I am wanting to start building high power audio amps. Of course, I realize the need to have a good, reliable bench power supply. I priced some and they are pretty outrageous. So, I figured I'd get a schematic and just build it first. The problem I am facing is that all the amp designs I have require higher voltages and the power supply schematics I find are all relatively low. What kind of sense does that make? I have one amp that needs plus and minus 85 volts at nearly 3 amps, yet I haven't found anything close to that. Granted, in these forums, I found higher voltage designs by some of the members, but they are all below 1 amp. Am I just looking for something that doesn't exist?
For testing just cobble one together from a few transformers rectifiers filter capacitors and a variac 2 35 0 35V 300 Va toroids wired in series with a 35A 800V bridge rectifier and 10000uf 100V caps for filtering will give 0 - 100V at 3A per rail more than enough to kill you (200V rail to rail) An option I have used before is to use a 52V transformer with a voltage doubler to get +/- 75V.
4 24V and 2 12V (low voltage lighting)transformers in series is an easy option if scrounging because 24V transformers are very common in industry The power levels you mention are not large, if you have a friendly electrical equipment recycler nearby all sorts of high powered goodies can be found just keep an open mind because there are many ways to get the required voltage. Like I mentioned before the voltages can be lethal and short circuits start to make loud bangs and lots of heat, DIN rail mounted MCB's are a cheap and effective way to protect things. There is not a lot of difference between low and high powered single phase power supplies only the component ratings and values change and the protection circuitry gets more sophisticated due to the greater value of the protected equipment. If you don't really understand power supplies then DF86's advice is applicable.
4 24V and 2 12V (low voltage lighting)transformers in series is an easy option if scrounging because 24V transformers are very common in industry The power levels you mention are not large, if you have a friendly electrical equipment recycler nearby all sorts of high powered goodies can be found just keep an open mind because there are many ways to get the required voltage. Like I mentioned before the voltages can be lethal and short circuits start to make loud bangs and lots of heat, DIN rail mounted MCB's are a cheap and effective way to protect things. There is not a lot of difference between low and high powered single phase power supplies only the component ratings and values change and the protection circuitry gets more sophisticated due to the greater value of the protected equipment. If you don't really understand power supplies then DF86's advice is applicable.
It's not necessarily that I do not understand power supplies, but rather I am transitioning from a technician standpoint to a design, develop and build standpoint, which is a bit different. I work at a steel plant, so I'm used to working with large voltages (118kv coming in, 38.8kv in half the plant before bring it down to 3 phase 480vac), but when it comes to power supplies, we just buy them and install them and completely replace them if they go bad. At work, I don't have to do component level troubleshooting, so I'm having to refresh my memory on this.
I do thank you both for your feedback though, and I will take your advice in starting with a lower voltage first. Any other advice or pointers would be happily accepted.
I do thank you both for your feedback though, and I will take your advice in starting with a lower voltage first. Any other advice or pointers would be happily accepted.
well having done a bit of research on this subject myself, i'm interested in the replies, burt I can tell you from my research I founf it cheaper to just get a used high quality Lab supply. the sort of voltage and low noise performance needed for a good bench lab supply will require a decent amount of effort, money and time. by the end of it you may end up spending more than buying a used or surplus supply.
its still appealing as project and you may of course build in features that you usually dont find and its a valuable experience, just saying be prepared to lose out monetarily, especially if looking towards the pointy end of parts quality and buying in small quantity, removing the economy of scale that commercial companies have. high end DIY is rarely about saving money is all i'm saying
its still appealing as project and you may of course build in features that you usually dont find and its a valuable experience, just saying be prepared to lose out monetarily, especially if looking towards the pointy end of parts quality and buying in small quantity, removing the economy of scale that commercial companies have. high end DIY is rarely about saving money is all i'm saying
Last edited:
Bench supplies to power amps at full tilt are more expensive and more massive than the most expensive amp you will build - because they are basically even bigger power amps in themselves with more features thrown in.
Buy a basic used twin 2-30V supply with current limiting and run up your amps on that first. It will run well enough to catch the bugs and save parts from smoking if you set the current limiter with anything more than minor care. You should be able to find a suitable supply for less than 100 dollars easy.
Buy a basic used twin 2-30V supply with current limiting and run up your amps on that first. It will run well enough to catch the bugs and save parts from smoking if you set the current limiter with anything more than minor care. You should be able to find a suitable supply for less than 100 dollars easy.
+/-85 V, 3 A? For an amplifier? Into what load? 32 ohm? +/-85 V rails in an amplifier driving into 8 ohms is more likely to require over 10 A. Anyway...
I second the reply of Metalsculptor. Get a 500 VA (or larger) variac (aka autotransformer). They can be had on eBay for slightly above scrap value. Buy a new one as buying used and fixing it will likely cost more than a new cheapie. Hook this up to a transformer with the needed bridge rectifiers and reservoir caps. Now you have a variable supply. The only catch is that you don't have a current limiter, though that could be implemented.
I doubt you'll find a lab supply kit that'll meet your needs. But if your goal is to learn how power supplies work, you could build a smaller one, get the knowledge, and transfer this knowledge into your own design of a higher voltage supply. You can get service manuals for many of the HP power supplies directly from HP (now Agilent). Just go to Agilent's website and download them. They can be a good source of inspiration for your own designs.
I have built test equipment (including many power supplies) since I was in 5th grade. But these days, every time I look at building my own gear, I can find a used piece of gear on eBay for a fraction of what it would cost me to build it myself. And I'm only counting the materials cost. Price out the transformer, heat sink, chassis, power transistors, and electrolytics for your project. Then add $100 for misc. discretes, connectors, knobs, meters, etc. That's probably a fair estimate. Makes a $200 used supply sound downright cheap... It's more fun building your own, though.
~Tom
I second the reply of Metalsculptor. Get a 500 VA (or larger) variac (aka autotransformer). They can be had on eBay for slightly above scrap value. Buy a new one as buying used and fixing it will likely cost more than a new cheapie. Hook this up to a transformer with the needed bridge rectifiers and reservoir caps. Now you have a variable supply. The only catch is that you don't have a current limiter, though that could be implemented.
I doubt you'll find a lab supply kit that'll meet your needs. But if your goal is to learn how power supplies work, you could build a smaller one, get the knowledge, and transfer this knowledge into your own design of a higher voltage supply. You can get service manuals for many of the HP power supplies directly from HP (now Agilent). Just go to Agilent's website and download them. They can be a good source of inspiration for your own designs.
I have built test equipment (including many power supplies) since I was in 5th grade. But these days, every time I look at building my own gear, I can find a used piece of gear on eBay for a fraction of what it would cost me to build it myself. And I'm only counting the materials cost. Price out the transformer, heat sink, chassis, power transistors, and electrolytics for your project. Then add $100 for misc. discretes, connectors, knobs, meters, etc. That's probably a fair estimate. Makes a $200 used supply sound downright cheap... It's more fun building your own, though.
~Tom
If you are a technician in a steel works, look around the plant for redundant equipment, 24V AC or DC PSU's abound in everything from motor starters to control panels, just avoid the switchmode ones. I would agree that buying a used power supply could be cheaper in the US, in Australia they are scarce. As for a current limiter use a double pole switch to insert a few suitable value high wattage resistors in series for initial testing. Just to give an idea of the range of transformers to be found in old equipment the last 3 transformers I removed from redundant equipment were 240/52V 0 52V 10A 240/ 52/48/42/36V 25A and 400/10V 7000A One piece of equipment had 15 100V 10000uF filter capacitors all with a nice 28A ripple current rating.
550W into 4r0 is equivalent to 66.3Vpk and 16.6Apk driving a sinewave signal into a 4r0 resistor.
For real speaker loading the amplifier should be able to pass upto 50Apk on very short term transients while holding the peak voltage up near that 66Vpk figure.
You cannot get that current demand from the transformer nor from a regulator nor an SMPS. The peak current is supplied from the smoothing capacitors and decoupling.
The bench supply will rarely if ever be used to "test" a real high power amplifier driving a real reactive speaker to the maximum current demand the speaker may have on exceptional musical transients. There is no point in designing a bench supply to meet that need, ever.
Andrew,
I wasn't planning on testing the circuit for functionality with an actual speak load; I planned on putting a signal from a signal generator on the input and a dummy load on the output and compare the input and output with a scope. I would think I would still need a bench supply that could output the required voltage and current demands of the circuit design, even if not into a reactive speaker load. The test would be to verify correct circuit fabrication more so than achievement of optimal design.
I wasn't planning on testing the circuit for functionality with an actual speak load; I planned on putting a signal from a signal generator on the input and a dummy load on the output and compare the input and output with a scope. I would think I would still need a bench supply that could output the required voltage and current demands of the circuit design, even if not into a reactive speaker load. The test would be to verify correct circuit fabrication more so than achievement of optimal design.
Hi,
then don't specify 3A for testing purposes.
0-30V & 3A is a reasonable spec for a bench supply. Two channels with both voltage and current metering on both channels (4meters) and adjustable current limiting.
A high voltage supply would be more likely specified as 20-100V & 1A
If you were working with valve circuits you might specify 100-400V & 300mA
then don't specify 3A for testing purposes.
0-30V & 3A is a reasonable spec for a bench supply. Two channels with both voltage and current metering on both channels (4meters) and adjustable current limiting.
A high voltage supply would be more likely specified as 20-100V & 1A
If you were working with valve circuits you might specify 100-400V & 300mA
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.